Linh is a health data scientist offering a blend of diverse expertise in clinical medicine and population health, study design and quantative data analysis.
Contact Linh
A data scientist: I love playing with data, working with stakeholders, and making better decisions with data.
An epidemiologist: My research interests lie in not only predictive modeling but also causal inferences with proper study designs.
A health professional: Although no longer working in clinical setting, my passion doesn't change - only shifts from the health of individual patients to that of wider populations.
About MeThis work fits a compartment model to the U.S. tuberculosis epidemic, using a Bayesian framework and written in Julia.
Github RepoThis application allows user to forecast the trend of country-level COVID-19 pandemic (pre-loaded) or other diseases (user-uploaded data), using phenomenological growth models. The application is built with Python and Shiny Dashboards.
Launch ApplicationThis application summarises key information of the COVID-19 pandemic at the country and global levels.
Launch ApplicationThis project looks for underlying homogeneous subgroups of frailty and investigate the progression of frailty over time in the elderly, by using latent transition analysis (LTA). The project uses LatentGOLD for LTA analysis, and R for data wrangling, descriptive analysis and visualization.
Github RepoThis project uses mixed-effects beta regression (since the outcome (frailty index) is bounded in [0, 1]) in APC framework to investigate frailty trajectories of different birth cohorts over time, and makes comparison between 9 European countries.The project uses R for analysis.
Github RepoThis project analyzes excess mortality rates due to influenza in Spain. The project applies the Serfling regression (Poisson and Negative Binomial Regression with seasonal components) to estimate excess deaths accountable to influenza outbreaks. The method and code can be applied to other disease outbreaks/pandemics. R is used for this project; SAS macro is also available.
Github RepoThis project utilize data from the Health and Retirement Survey (HRS) to predict the 2-year fall risk of Americans aged 65 or older, using logistic regression (LR), k-nearest neighbors (KNN), Support Vector Machine (SVM), random forest (RF), multilayer perceptron (MLP) with two hidden layers of size (20, 20). This project uses R and Python for analysis.
Read ReportOver six (6) years of research experience in designing and implementing studies and analyzing quantitative observational and randomized trial data. Well-versed in statistical and mathematical modeling, including linear and non-linear mixed models, multivariate analyses, causal inferences, and predictive modeling.
Linkedin Download My Resume